DESCRIPTION The overall objective of the proposed FIRCA project is the development of improved models for structure-based prediction of ionization constants of titratable groups in proteins. The success of this project would contribute directly to out understanding of enzyme mechanisms, protein stability and molecular recognition, and ultimately might be extremely helpful in rational theoretical drug design. Current methodology, developed within the parent NIH grant, takes into account changes in charges of all the atoms of titratable groups, together with explicit changes in the dielectric boundary of the group associated with the ionization process. In the proposed work, the model will be expanded to consider the orientational equilibria of polar hydrogens and to explicitly include water molecules into simulations. Work will also be undertaken to simulate titration behavior of a large number of structures. The methodology will be based on continuum dielectric electrostatics and the finite-difference Poisson-Boltzmann approach. It is sufficiently fast that a large number of proteins can be investigated within a short time using only a standard computer work station. This will provide the necessary statistics to evaluate the validity and predictive power of the methodology and facilitate the discovery of desired improvements and corrections for parameters and modeling.